According to the US Department of Labor, the cumulative cost of wages, productivity losses, and administrative delays resulting from workplace injuries amounts to $161.5 billion annually. Among the causal factors are the improper use of PPE—with, for instance, the construction, mining, and manufacturing industries accounting for 40% of eye injuries in workplaces alone—the lack of comprehensive monitoring mechanisms, and the lack of insight into risk patterns.
With the nature of safety risks varying across industries—ranging from hazardous chemicals (pharma) and equipment failure (manufacturing) to spills and leaks (O&G), for example—the need to standardize safety procedures, implement centralized monitoring, and automate pattern-detection is imperative. Amidst sea changes across industries, we find that businesses are now turning to existing and new AI technologies to deploy safety strategies for employees and customers alike. These solutions have been implemented across a variety of functions, broadly classified as follows:
In fact, according to research, 6.2 billion hours of worker productivity will be recovered by 2021 using AI augmentation, with $2.9 billion in business value simultaneously generated. Let us look at how this is playing out at the ground level, and how businesses can leverage the latest technology to usher in a new age of workplace safety.
While PPE use following the pandemic has become commonplace, there remain several areas of operation, across industries including manufacturing and construction, where non-compliance with PPE-use guidelines have resulted in severe damages and losses. According to the U.S. Bureau of Labor Statistics, only approximately 1% of workers with face injuries were found to be wearing face shields and only 23% of workers sustaining foot injuries were using safety footwear. While the situation has improved considerably since the implementation of the Occupational Safety and Health (OSH) Act in the United States in 1970, the room for more improvements and streamlined processes is immense.
Further, with the use of PPE including gloves, safety hats, safety footwear and eyewear, masks, reflective clothing, harnesses, and visors varying from function to function, manual monitoring becomes a cumbersome and highly time-consuming task. The use of AI-driven computer vision technology here can not only tailor monitoring to various employee groups, for example distinguishing between the PPE required for doctors and that required for factory staff, but also allow real-time alerts through automated notification systems. AI & ML algorithms can measure images and real-time video feeds across networks against safety protocol guidelines to flag off instances of non-compliance, risks, and more.
Video analytics can be implemented at several touchpoints to ensure compliance, with automatic notifications sent to managers/supervisors in case of issues. Further, AI can also monitor work environments through heat, audio, and IoT sensors, highlighting increased levels of chemicals/heat in the environment as well as turning off machinery/barring access if an employee is not wearing appropriate PPE. Following data collated over a longer period, safety procedures can be better designed by accounting for relevant risk areas, and training sessions reinforcing PPE use among employees can also be carried out to better effect.
With the imminent return to workplaces in certain parts of the globe, the need for appropriate distancing and screening remains as important as it was during the pandemic. While we saw automated thermal screening adopted by several organizations, the scope for this technology to be used more widely remains, with temperatures checked throughout the day and any issues immediately reported, to enable rapid isolation, checks, and testing. In addition to thermal detection, social distancing and proximity awareness, PPE use, physical symptoms monitoring, and so on should also be implemented for the short term.
Contact-tracing through mobile applications and wearables can help boost screening and distancing measures even further. While computer vision-enabled cameras can measure the distance between employees, instant alerts also need to be set up in case of breaches. What becomes important here is a means of tracking groups of employees who interact with each other frequently; using these interactions to disable or customize distancing notifications and enabling quick response in case a particular employee tests positive. AI-linked devices including wearables and smart ID cards can enable monitoring for hundreds of individuals across geographies, with real-time updates on location and health status. Apart from helping improve workflows, in terms of mapping the locations of materials, tools, and workers and guiding employees to the right locations, wearables can be used to both deliver instant notifications as well as inform employees about crowding/capacity in shared workplace locations. Following this, shifts can be better planned to ensure optimal distancing, reduce conflicts, and streamline access to resources.
A major revolution across industries dealing with heavy machinery, complex production lines, and so on has been the move from reactive to predictive maintenance. With the proliferation of IoT and sensor technology, as well as the seamless integration of AI & ML in data collation and analysis, businesses can now extend asset lifecycles, undertake maintenance measures to prevent downtime, and get foreknowledge on the risks of certain equipment breaking down or wearing out.
AI can be used not only to reduce downtime and related costs but also to augment employee safety, combining predictive maintenance data with employee shift, workflow, and physical movement data. Real-time analytics can enable comprehensive monitoring, and the resultant information can be made available on intuitive dashboards across sites, on hand-held devices, and wearables to facilitate quick decision-making and response.
According to research, falls (33.5%) and objects (11.1%) are among the major causes of construction injuries in the United States. Further, the U.S. Bureau of Labor Statistics identifies transportation incidents among the topmost hazards across workplaces, with 39.6% of them being fatal. Considering all these factors, here are a few examples of how AI can help augment on-site safety:
In the case of workplaces where manual labor can be hazardous, technology can take over from human beings, reducing safety risks and increasing time to speed through automation. For instance, drones can be used to lift heavy loads and transport equipment from one place to another, results in time savings and safety. Moreover, they can also be used to gain aerial views of sites, which can aid site planning and decision-making as well as safety processes.
Moreover, collaborative robots, or co-bots, have come to prominence in recent times, aiding human labor in factories and other workplaces. In addition to being able to monitor employee safety, we have seen that such robots can also take on more hazardous tasks, including precision welding, working with dangerous chemicals, installing large objects, transporting heavy loads, and so on, which can help ensure greater employee safety in the longer run.
In addition to screening employees for health-related indicators, the need to monitor their work environments and enable safety also becomes important. For instance, technologies to screen people entering buildings can be deployed to ensure that there are no intrusions and that only authorized people can access certain areas of the workspace. Such AI-driven technology can extend to any premises, screening vehicles and entrants to ensure that only authorized personnel have access. Computer vision technologies can be utilized here to verify the identity of people against internal records. Following issues, notifications can immediately be sent to security teams to enable quick responses.
Such measures also extend to the virtual space, where increasing cyberattacks can pose a major threat to employees’ personal information. AI-driven cybersecurity solutions can undertake a wide variety of screening and detection tasks, from login authentication and threat hunting to vulnerability management and real-time fraud detection. Virtual intrusions, attacks, abnormal online behaviors, and so on can therefore be immediately detected, ensuring the safety of personal information and IP.
With the severe impacts of safety hazards in the workplace, the need to ensure more optimal workplace safety measures has become a priority, especially following the pandemic. While the pervasive use of cloud and AI-driven technologies has made safety solutions widely available across industries, the question of the ethical use of monitoring mechanism and personal information may be called into question here. Businesses must strike a balance by demonstrating transparency with employees as well as employing methods of anonymizing data, which will be crucial to building a safe and reliable workplace. With AI transforming the parameters of safety in workplaces, greater employee engagement and wellbeing is now within reach for businesses.